Anthropic secured US government permission to redeploy its Mythos 5 cybersecurity AI model in a partial reversal of a recent export control directive. Over 100 companies and government agencies are now authorized to use the system, including their non-American employees.
The move comes two weeks after an export control directive forced Anthropic to disable access to both Claude Fable 5 and Claude Mythos 5 for all customers. The new deal represents a limited victory for the AI firm, allowing it to restore service to select organizations while leaving other restrictions in place.
According to TechCrunch, the authorization covers more than 100 US entities, with Engadget confirming that Anthropic is restoring access to select organizations. The directive was issued by the Trump administration, according to multiple reports.
The decision primarily benefits US-based defense and cybersecurity operations that rely on Mythos for threat detection. Non-American employees of authorized organizations can also use the system, though broader international access remains blocked.
Critics argue the partial lift does not go far enough, leaving Anthropic's global customers locked out and creating an uneven playing field for international partners who depend on the same AI tools.